Laboratory Coolants: Choosing Water or Glycol for a Recirculating Loop
Share
A recirculating bath can show the right setpoint while the equipment it serves receives too little coolant. A fluid that remains liquid in the reservoir may be too viscous at startup, may form slush at a colder surface, or may attack a seal hidden inside a quick connector. The right choice is the fluid that can carry heat through the whole loop at the required temperature and flow, within the limits of every part it touches.
Learning objectives: Use a fluid chart to check freeze margin; explain why a stronger glycol mixture can reduce flow; distinguish fluid approval from a catalog temperature range; and plan a supervised checkout that catches problems before a sample or heat source is introduced.
What a coolant has to do
In an external circulation loop, the pump sends liquid from the temperature control unit through a load, such as a glass jacket or a metal heat exchanger. The return carries absorbed heat back to the unit. The refrigeration circuit then has to move that heat out of the circulating liquid and reject it to the room or a facility cooling service. A controller display typically reflects a particular sensor location; it does not by itself establish the load temperature, actual flow, or steady cooling capacity. During startup or a disturbance, these can tell very different stories. For the separate question of how vapor and coolant travel through glassware, see the laboratory condenser guide.
Specific heat describes how much heat a unit mass of liquid carries per degree of temperature rise. Density tells you how much mass is delivered at a given volume flow. Thermal conductivity matters at the wall, where energy must pass between the process and the fluid. Dynamic viscosity describes resistance to shear. In the same circuit, thicker fluid generally requires a greater pressure difference to maintain the same flow. The change you actually see depends on the pump curve, the restrictions in the loop and where the flow is measured. Sun and Teja measured density, viscosity and thermal conductivity for several aqueous glycol systems and found that the properties depend on composition and temperature; their published measurement range does not establish subzero values for a commercial coolant. Use the fluid supplier’s data at your working conditions. [1, 2]
Water often makes an efficient coolant where the manual permits it: it has high heat capacity and low viscosity compared with many aqueous glycol blends. Adding an approved glycol can lower the temperature at which ice first forms, but it changes other properties at the same time. Increasing concentration may buy a larger freeze margin while making cold circulation more difficult. A pump’s advertised maximum flow is commonly reported at particular test conditions and can be far above the flow delivered through a narrow jacket, long hoses or a restrictive quick connector. The operating point is where the pump and the assembled circuit meet.
Figure 1. The coolant travels around the loop; the process stays in its own passage. Compare the indicated bath temperature with readings near the load if suitable measurement ports are available. The schematic omits safety devices and is not a plumbing drawing.
Fluid and equipment compatibility
Gather the exact unit’s manual, the load’s limits, and the candidate fluid’s current technical data and safety data sheet. Map the wetted path all the way from the reservoir through the pump, heat exchanger, hoses and couplings to the load and back. Include parts that are easy to overlook: an O-ring behind a quick connector, the plastic body of a flow sensor, a metal insert in a fitting or the adhesive in a hose assembly. Enter the minimum and maximum exposure temperatures, expected pressure including startup or blocked-flow conditions, and the planned concentration. Check component limits for those combined conditions. A fluid listed as compatible with a tank metal at room temperature is no answer about the pump seal, hose or glass jacket at the cold end of the range.
The 6 L laboratory chiller circulator illustrates how to read a catalog entry. The listing gives a nominal 6 L capacity, a stated −20 to 100 °C range, internal and external circulation, and a 304 stainless steel tank and worktop. It does not give the complete wetted materials list, approved fluid chart, loaded cooling curve or delivered pump flow. The capacity is not a fill instruction, and the endpoints are not permission to use the same fluid at both extremes. For a glass load, the 100 mm borosilicate reflux condenser lists a 24/40 process joint. That identifies a glass connection; it does not specify the coolant fitting dimensions, jacket pressure limit or acceptable thermal step. These two listings are examples of what to verify, not a documented pairing. Compare similar loads in the laboratory glassware collection using their own documentation.
- Water, when the manual allows it.
Check the approved water grade, conductivity or resistivity limits, and minimum temperature. Tap water can deposit scale as water evaporates; clear-looking water can still have dissolved minerals. Ultrapure deionized water is also a poor default: some equipment specifically requires it, while other units restrict it because of their materials or water quality requirements. Thermo Scientific, for example, specifies a resistivity limit for water on its Accel series. That limit belongs to the equipment in that manual. [1, 3]
- Approved inhibited glycol and water.
Verify whether the equipment permits propylene glycol, ethylene glycol or a named premix, and in what concentration range. Ethylene glycol presents a serious ingestion hazard [8]; propylene glycol generally has lower acute oral toxicity, but its additives and any contamination still require SDS and waste review. Inhibitors are formulation specific, and using too little or too much concentrate can move the fluid outside its approved range. Do not swap glycol types, add a biocide, or mix brands to correct problems unless the manufacturers permit it. [1, 4, 5]
- Other specialty fluids.
Use only if the unit and application explicitly approve them. A silicone-based thermal fluid may have a different seal compatibility profile and handling requirements. Automotive antifreeze, brine and alcohol are not substitutes chosen by freeze point alone. Unknown inhibitors, salt corrosion, or flammable vapor can introduce hazards the equipment was not designed to control. Follow the exact equipment documentation and site chemical review rather than trying an improvised blend. [1, 3]
Worked example: check the freeze margin
Suppose a materials teaching lab needs an external metal test block held near −5 °C. Its equipment documentation and a review of startup and idle conditions give an estimated coldest relevant liquid temperature of −8 °C, including the local temperature near the refrigeration surface. Assume this hypothetical setup requires a 4 K margin below that coldest condition. The target freezing point is therefore −8 °C − 4 K = −12 °C or colder. A temperature difference of 4 K has the same magnitude as a difference of 4 Celsius degrees. Neither the setpoint nor this illustrative margin replaces the installed unit’s instructions.
Dow’s DOWFROST data report a typical freezing point of −12.7 °C at 30.0% propylene glycol by mass. The same row gives 29.4% propylene glycol by volume and 30.7% DOWFROST concentrate by volume. These columns describe different quantities: the commercial concentrate contains additives and is not pure glycol, and mass fraction is not volume fraction. The freeze margin using the stated values is −8 − (−12.7) = 4.7 K. That clears the assumed 4 K criterion on paper, with just 0.7 K beyond it. If the coldest condition instead reaches −10 °C, the margin falls to 2.7 K and the candidate fails that screen. Concentration error and temperature uncertainty could also consume a small apparent excess. This is a comparison of typical fluid values, not a mixing prescription or approval for the catalog circulator. [4]
Freeze protection and burst protection answer different questions. A chart entry for burst protection concerns whether expansion is likely to rupture equipment under specified conditions; slush may already have formed and stopped circulation. A working loop needs pumpable liquid and an operating viscosity the pump can handle. As a separate illustration, Dow data for a 30% by volume propylene glycol mixture give about 4.51 mPa·s at 10 °C and 1.63 mPa·s at 40 °C. The ratio is about 2.8. In a straight round tube with fully developed laminar flow and fixed geometry and volume flow, the frictional pressure drop would change by the same factor. This is an illustration of sensitivity to temperature, not a prediction of the total drop through a real jacket or the performance of either catalog item. The fittings, pump and flow regime may change that answer. Neither published point should be extrapolated to −8 °C. [1, 2, 4]
Prepare and check a recirculating loop
1. Define the acceptance checks before filling. Record the exact fluid and its allowed concentration range, anticipated minimum temperature, freeze margin, acceptable flow at the load, and any maximum allowable pressure of the weakest component. Verify the chiller’s rated cooling capacity at the required temperature and ambient conditions; a refrigeration capacity given at a warmer setpoint is not evidence of the same performance in a cold loop. Identify the manufacturer’s fill marks, priming instructions and alarm response. If a needed rating or approved fluid is unknown, obtain it before use rather than selecting a trial temperature.
2. Inspect and prepare at a safe temperature. Follow the SDS and wear chemical splash goggles, suitable gloves and the protective clothing required by the lab. Inspect glass for chips, cracks and damaged sidearms. Use hoses and retainers that fit the actual fittings and are rated for the fluid, temperature and pressure. Support heavy hoses independently so they do not lever against the glass. Prevent kinks, leave room for normal motion, and keep electrical connections away from a possible leak. Set up containment and route condensate away from electrical equipment: a hose below the room’s dew point can drip even when its connections are tight. Do not put an improvised valve or flowmeter in the line; added restriction matters, especially for glass. [3, 6]
3. Fill with an identified fluid. Use an approved premix or follow the fluid maker’s instructions for dilution and water quality; record whether the stated concentration refers to glycol or concentrate and whether it is by mass or volume. Keep unknown residual fluid out of the new charge by following the manufacturer’s change and rinse instructions. Fill to the designated mark and prime as directed. Never run a pump dry to clear air. After the fluid is uniform, use a compatible concentration test with the correct temperature correction and formulation chart, if the supplier calls for one. Record the result and test method next to the batch details. [3, 4, 5]
4. Establish a baseline with the process idle. Under supervision, circulate without process heat or a sample and check the indicated level after the external loop fills. Look for drips at connectors, bubbles, irregular pump sound and interrupted return. Record flow using approved instrumentation or the documented equipment check. Avoid disconnecting a pressurized return for a bucket test. As the unit approaches the approved target, record the supply and return temperatures and the flow after stabilization; trace weak flow while conditions are still safe. A clean idle check is a prerequisite for operation, not proof of loaded performance. Only then apply the approved load and verify that flow, pressure and temperature stay within limits.
5. Shut down and manage spent liquid. Follow the shutdown sequence for the load and temperature control unit. Let the assembly reach a safe service temperature and isolate power and pressure as required before disconnecting hoses. Collect spent glycol and potentially contaminated rinse liquid in compatible, labeled containers. Report what was actually in the loop, including any suspected process contamination, to your lab’s waste program. Do not discharge it to a sink or storm drain without explicit site authorization. The University of Michigan, for example, directs collection of glycol solutions through its campus waste program. Treat spills as slip and exposure hazards and follow local procedures. [6, 7]
Troubleshooting and stop criteria
- Flow is steady when warm but falls during cooldown.
Pause the load. Compare viscosity at the actual cold temperature with pump limits and check for slush, a collapsed or kinked hose, or a newly restricted filter. Where a device gives a low-flow alarm, follow its reset instructions only after correcting the cause. Raising pump pressure blindly can transfer the problem to a fragile jacket.
- The display is steady but the load drifts.
Measure closer to the load with a documented sensor arrangement. Compare supply temperature, return temperature, flow and ambient conditions with the idle baseline. A warm return with normal circulation can suggest a larger heat load; inadequate flow or a setpoint that cannot be maintained under load calls for a different investigation. An infrared reading on a wet or reflective surface cannot establish fluid temperature; see the guide to infrared and probe thermometers.
- The fluid changes color, turns cloudy or the level drops repeatedly.
Check for leakage, contamination, concentration change or material deterioration. Do not infer glycol concentration or inhibitor condition from color. If the process and coolant might be mixing across a damaged wall, stop and isolate the equipment under the lab procedure, mark it out of service and treat the liquid as potentially contaminated.
Frequently asked questions
Should I top off with water or premix?
Find out what escaped first. Evaporation often removes mostly water; a liquid leak removes the blend. Use the manufacturer’s instructions and a concentration check to decide, then investigate repeated loss. Automatically adding premix does not preserve concentration in a bath that mainly loses water.
Can I use a sugar refractometer to read percent glycol directly?
A Brix scale measures refractive response as a sugar equivalent. It is not automatically a glycol percentage. Use a scale calibrated for the intended glycol or a model and correction chart the fluid supplier approves. A sample with mixed fluids or contamination may invalidate the conversion. [4, 5]
Can one coolant cover every seasonal or process setting?
Only if the exact formulation and all wetted components are approved over that entire range and the assembled loop delivers acceptable flow and capacity. If different fluids are needed, plan the changeover and waste collection rather than layering one on top of the other.
Technical references
- Thermo Fisher Scientific: Proper Fluid Selection and Maintenance — water grades, corrosion and fluid properties.
- Sun and Teja, Journal of Chemical & Engineering Data (2004), DOI 10.1021/je049960h — measured transport properties of aqueous glycols; NIST ThermoML record.
- Thermo Scientific: Accel Series operating manual — model-specific fluid approval, cold-surface margin and wetted materials.
- Dow, DOWFROST typical freezing points and thermophysical properties — formulation-specific examples.
- Dow: Heat Transfer Fluids Calculators and engineering guides — formulation selection and operating guidance.
- University of Vermont EHS: Water-Cooled Equipment — hose security, pressure and spill concerns.
- University of Michigan EHS: Open and Closed Loop Wastewater — institutional guidance for collecting glycol waste.
- ATSDR: Ethylene Glycol ToxFAQs — ingestion hazard.
This tutorial is for laboratory education. It does not set operating limits for a particular unit or establish a validated analytical, clinical or regulatory method.